Legislative Elections 6/27/96

[Steve Eppley]

Demorep1 wrote:
>Legislative Elections with Combined Choices Tie Breaker
-snip-
>Due to the major complexity of the following, it would be much
>easier just to repeatedly drop the candidate with the lowest number
>of first choice votes (or first choice plus transferred votes) until
>only N candidates remain (i.e. no head to head math or quotas). 
-snip-

But if votes in excess of the win-quota aren't transferred, this
will underrepresent the supporters of the more popular candidates and 
overrepresent the supporters of the less popular candidates.  It 
could easily produce minority rule:

Example: 3 seats to be filled, 5 candidates from two parties running.
  98: L1 > L2 > L3
   1: R1 > R2
   1: R2 > R1
To me it's clear that L1, L2, and L3 should win the three seats. 
But by avoiding the complexity of overquota transfers, you'd elect
L1, R1, and R2.  The party of two elect themselves into control.

The complexity of transferring the excess as in STV must not be avoided.  

In the example above, if it takes 26 (hare quota) votes to win a seat, 
then L1 will only consume 26 votes, leaving 72: L2>L3. 
Then L2 will win a seat, consuming 26 more, leaving 46: L3. 
Then L3 will win the final seat.

(Even if a Disapproval step is added, I could still design a more
fragmented example which illustrates the minority rule flaw.)

---Steve     (Steve Eppley    seppley at alumni.caltech.edu)